spine clinical article J Neurosurg Spine 23:518–525, 2015

Surgical outcomes of 156 spinal accessory injuries caused by biopsy procedures

Sang Hyun Park, MD, PhD,1 Yoshua Esquenazi, MD,2 David G. Kline, MD,3 and Daniel H. Kim, MD2

1Department of Anesthesiology and Pain Medicine, Jeju National University Medical School, Jeju, Korea; 2Department of Neurosurgery, The University of Texas Health Science Center at Houston Medical School, Houston, Texas; and 3Department of Neurosurgery, Louisiana State University Health Sciences Center, New Orleans, Louisiana

Object Iatrogenic injuries to the spinal accessory nerve (SAN) are not uncommon during lymph node biopsy of the posterior cervical triangle (PCT). In this study, the authors review the operative techniques and surgical outcomes of 156 surgical repairs of the SAN following iatrogenic injury during lymph node biopsy procedures. Methods This retrospective study examines the authors’ clinical and surgical experience with 156 patients with SAN injury between 1980 and 2012. All patients suffered iatrogenic SAN injuries during lymph node biopsy, with the vast majority (154/156, 98.7%) occurring in Zone I of the PCT. Surgery was performed on the basis of anatomical and electro- physiological findings at the time of the operation. The mean follow-up period was 24 months (range 8–44 months). Results Of the 123 patients who underwent graft or suture repair, 107 patients (87%) improved to Grade 3 functional- ity or higher using the Louisiana State University Health Science Center (LSUHSC) grading system. Neurolysis was performed in 29 patients (19%) when the nerve was found in continuity with recordable nerve action potential (NAP) across the lesion. More than 95% of patients treated by neurolysis with positive NAP recordings recovered to LSUHSC Grade 3 or higher. Forty-one patients (26%) underwent end-to-end repair, while 82 patients (53%) underwent graft repair, and Grade 3 or higher recovery was assessed for 90% and 85% of these patients, respectively. The average graft length used was 3.81 cm. Neurotization was performed in 4 patients, 2 of whom recovered to Grade 2 and 3, respectively. Conclusions SAN injuries present challenges for surgical exploration and repair because of the nerve’s size and lo- cation in the PCT. However, through proper and timely intervention, patients with diminished or absent function achieved favorable functional outcomes. Surgeons performing lymph node biopsy procedures in Zone I of the PCT should be aware of the potential risk of injury to the SAN. http://thejns.org/doi/abs/10.3171/2014.12.SPINE14968 Key Words lymph node biopsy; nerve action potential; nerve injury; nerve repair; spinal accessory nerve; iatrogenic injury

he posterior cervical triangle (PCT) is an ideal ac- the muscle and leads to weakness of the shoul- cess point for many surgeons to perform a lymph der in abduction, dropping of the shoulder, and winging of node biopsy. The lymph nodes of the are su- the with pain and stiffness in the shoulder girdle. perficialT in this zone and relatively simple to extract com- Spontaneous recovery after an SAN injury is reported in- pared with the nodes in other areas of the neck, which are frequently, and if the lesion is left untreated, it will lead to generally covered by abundant fibrous and fatty tissues. pain and functional deficit in 60%–90% of cases.1,2 While Due to the close relationship between the lymph nodes and the critical time for surgical repair after the injury remains the spinal accessory nerve (SAN), approximately 3% to unknown, favorable outcomes can be achieved following 10% of all lymph node biopsy procedures result in SAN early surgical management. This report presents 32 years injuries.1 SAN injury results in loss of motor function of of surgical experience with 156 SAN iatrogenic injuries

Abbreviations LSUHSC = Louisiana State University Health Science Center; NAP = nerve action potential; PCT = posterior cervical triangle; SAN = spinal accessory nerve; SCM = sternocleidomastoid muscle. submitted September 22, 2014. accepted December 31, 2014. include when citing Published online June 26, 2015; DOI: 10.3171/2014.12.SPINE14968. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

518 J Neurosurg Spine Volume 23 • October 2015 ©AANS, 2015

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC Spinal accessory nerve injuries caused by lymph node biopsy that were due to cervical lymph node biopsy and required when some portion of the arc of abduction movement was surgical intervention. We present the surgical anatomy possible with gravity eliminated; 3 (fair) when the arc of and clinical characteristics of the SAN and PCT, as well abduction against gravity was at 90° or better; 4 (good) as operations performed and their outcomes, to provide when strength of abduction against gravity and some re- management guidelines for SAN injuries. sistance were possible; and 5 (excellent) when strength of abduction was against moderate-to-full resistance (Table Methods 2). Each patient’s initial functional impairment and recov- ery at the time of each follow-up evaluation were assessed. Patient Population and Evaluation Grade 3 or better recovery was considered a favorable Between 1980 and 2012, two senior authors (D.G.K. functional outcome. and D.H.K.) surgically managed 156 patients with iatro- genic SAN lesions following cervical lymph node biopsy Anatomy of the SAN procedures without spontaneous recovery. Functional out- The SAN is only 1 to 2 mm in diameter and originates come after surgical repair of the SAN lesions in relation from both a cranial and spinal root.6,19 The cranial root to the type of surgical repair was assessed. There were 89 arises from the vagal nuclei and innervates some of the men and 67 women, and the mean follow-up period was muscles of the soft palate and larynx.9 The spinal root 24 months (range 8–44 months). The patients’ ages ranged arises from the spinal accessory nucleus within the C1–5 from 6 to 74 years (mean 38 years; Table 1). vertebrae and innervates the sternocleidomastoid muscle Each patient’s lymph node biopsy incision sites were (SCM) and trapezius muscles. Damage occurs more com- documented and noted on initial evaluation. Once an injury monly to the spinal portion of the SAN than to the cranial was diagnosed, the SAN lesions were monitored clinically portion. and by for several months for signs of The path of the SAN is variable, but generally travels early regeneration. Due to the nature of our practice, only in a line through Zones I and II of the PCT (Fig. 1). It ex- those patients who experienced unsuccessful conservative its the skull through the , and lies lateral management and for whom spontaneous regeneration did to the and immediately anterior to not occur were referred to us for evaluation. Therefore, pa- the transverse process of the atlas vertebra.9,14 The nerve tients who required surgical intervention made up the co- runs from the posterior or lateral border of the SCM be- hort for this study (n = 156); a control group was not avail- ginning near the muscle’s origin from the mastoid process able. Surgical exploration was performed on all patients, at an average length of 6.13 cm.17 Here the SAN is very which included intraoperative stimulation and nerve ac- superficial, lying in the deep cervical fascia embedded in tion potential (NAP) recordings to evaluate the degree of fibroadipose tissue and lymph nodes,19 and near the great physical continuity for each lesion to determine the need auricular nerve, which wraps around the lateral border of for repair by nerve grafts, end-to-end sutures, neurolysis, the SCM cephalad. or neurotization. The SAN and great auricular nerve form an important Before and after surgery, each patient underwent a landmark for surgeons.16 The cephalad portion of the SAN thorough clinical assessment of SAN motor function. has been identified to be 0.5 cm to 1.5 cm superior to the Functional assessment included measurement of active egress of the great auricular nerve at the posterior border abduction of the shoulder; manual testing of the trape- of the SCM.13 The nerve point, which is where the bundle zius muscle; and evaluation of the contour, function, and of sensory (C-2, C-3, and C-4) from the cervical strength of the trapezius muscle. The contraction of the plexus arises from beneath the posterior border of the trapezius and its ability to stabilize the scapula during a SCM, is also an important anatomical landmark for lo- complete arc of abduction were also assessed and com- cating the SAN.9 In the PCT, the SAN assumes a coiled pared with those of the normal side. S-shape when it is relaxed. The shape of the nerve allows The study used the Louisiana State University Health flexibility when the head is turned, and it straightens when Science Center (LSUHSC) grading system to document the degree of the injury for each patient as follows: a grade TABLE 2. Functional assessment grades of 0 (none) was given when muscle contraction was clini- cally absent; 1 (trace) when muscle contraction was clini- Grade* Evaluation Description cally evident but limited to a flicker of movement; 2 (poor) 0 None Muscle contraction was clinically absent 1 Trace Muscle contraction was clinically evident but lim- TABLE 1. Demographics and follow-up period for 156 surgical ited to a flicker of movement patients with SAN lesions due to cervical lymph node biopsy 2 Poor Some portion of the arc of abduction movement (1980–2012) was possible w/ gravity eliminated Variable Value 3 Fair Arc of abduction against gravity was at 90° or better Sex 4 Good Strength of abduction against gravity and some Males 89 resistance were possible Females 67 5 Excellent Strength of abduction was against moderate-to- Mean age in yrs (range) 38 (6–74) full resistance Mean follow-up in mos (range) 24 (8–44) * Based on the LSUHSC grading system.

J Neurosurg Spine Volume 23 • October 2015 519

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC S. H. Park et al.

Fig. 1. Illustration of the PCT divided into Zones I–III. The SAN is most susceptible to iatrogenic injury in Zones I and II; Zone III is the recommended cervical lymph node biopsy site. Copyright Daniel H. Kim. Published with permission. Figure is available in color online only. the shoulder is simultaneously lowered.6,7 By the time the the SAN in the PCT has been reported to range from 1.5 SAN runs through the posterior triangle, it is the only to 9 cm, with a mean length of 5.27 ± 1.52 cm and a depth source of motor supply to the trapezius. The general ex- of less than 0.4 cm.4,6 tracranial length of the SAN varies from 7 cm to 18.5 cm, with a mean (± SD) length of 12.02 ± 2.32 cm.4 Posterior Cervical Triangle Lymph Node Chain About 300 of the approximately 800 lymph nodes in the Posterior Cervical Triangle human body reside in the neck area, running roughly ver- After supplying the SCM on its deep surface, the SAN tically through its structures (Fig. 2). The size of a lymph emerges from its posterior border to enter the PCT.17 The node varies from 3 mm to 25 mm in diameter.20 The level PCT can be divided into 3 zones (Fig. 1), where Zones I II area of the neck, where the PCT is located, is an impor- and II represent areas in which the SAN is susceptible to injury. Cadaveric studies have reported variation in the course of the SAN across the posterior triangle.6 In 1 study of 25 cadavers, the nerve ran inferiorly and laterally to the anterior border of the trapezius in 74% of the cadavers, and inferiorly and laterally in a tortuous line in 26% of the cadavers.17 In 14 of the dissections, the nerve ran ante- riorly along the border of the trapezius toward the clavicle for several centimeters before entering the muscle.17 In a more recent study of 10 adult human cadavers, all of the SANs had a straight gross configuration proximal to the PCT and a tortuous appearance within the PCT.18 The SAN passes obliquely beneath the deep fascia on the levator scapulae, where the C-2 and C-3 vertebrae pro- vide neural contributions.19 The nerve lies superficial and close to in this area, and it is here that a SAN injury will result in major trapezius motor loss. The SAN supplies short branches to the trapezius along the superior and posterior border of the neck and longer branches to descend into the thoracic portion of the mus- Fig. 2. Posterior cervical lymph node chain (green) and course of the cle. At the edge of the trapezius, the SAN is approximately SAN. Copyright Daniel H. Kim. Published with permission. Figure is 2.7 ± 0.9 cm caudal to the cricoid cartilage.6 The length of available in color online only.

520 J Neurosurg Spine Volume 23 • October 2015

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC Spinal accessory nerve injuries caused by lymph node biopsy tant area for neck dissection because approximately 25% main pathology. For SAN injuries in the PCT, this can be a of cervical lymph nodes are in this area. These nodes have difficult maneuver because a mixture of scar and adipose a site-specific drainage pattern and the chain of nodes tissue often lies in the upper part of the PCT between the there communicate with the inferior and superior ends of posterior border of the SCM and the anterior border of the the jugular nodal chain by the transverse cervical node trapezius. Dissection under the operating microscope or group.15 As shown in Fig. 2, throughout its path in the PCT magnifying loupes may take considerable time as the sur- to the investing layer of the deep cervical fascia, the SAN geon distinguishes the and other sensory lies within fibrofatty tissue and courses superficially to a branches to the skin of the neck and periaural regions. chain of 4 to 23 lymph nodes.4,9,15,20 Afferents of the spinal The soft tissue and/or scar must be searched on the an- accessory nodes in this area originate from the occipital, terior border of the trapezius to find fascicles or branches postauricular, and suprascapular nodes. indicating motor entry points. This is especially important where a length of the SAN has been resected at the time Surgical Procedures of lymph node removal. The proximal nerve runs between Surgery was performed under general anesthesia with- the sternomastoid and cleidomastoid bellies of the SCM, out the use of paralytic agents, enabling identification of and a neuroma may be buried there in adipose and fibrous the proximal nerve stump with nerve stimulation. The in- tissue. The proximal and distal stumps are prepared us- cision for lymph node biopsy with resultant SAN palsy ing the standard method (Fig. 4 upper). If mobilization was extended or turned into a Z-shaped incision by add- of 2 proximal and distal stumps can be brought together ing a limb either superiorly or inferiorly. This extension without significant tension, end-to-end anastomosis can be allows proximal exposure of the SAN and distal exposure achieved with 8-0 nylon sutures (Fig. 4 lower). This proce- of its branches as they reach the trapezius muscle (Fig. dure is superior to nerve grafting when possible and espe- 3). The SAN usually exits between the lateral border of cially when the transection is relatively clean, because the the sternomastoid and cleidomastoid muscles just superior structure can be aligned. to the point where the auricular sensory branch runs over Nerve grafting is preferred when the SAN is lacerated. the SCM. This branch can be easily identified during the Likewise, transection requires grafting if the nerve ends initial stages of dissection because it curves around the cannot be brought together without tension. A sural nerve posterior border of the SCM. The proximal stump of the is preferred if nerve grafting is chosen, and it should be di- SAN is usually found approximately 1–2 cm superior to vided beneath the fascia in the lower leg.1 The donor graft this nerve branch. The proximal stump or branches should length should be chosen to allow ample movement of the be preserved during the procedure. The branches from the patient’s neck after surgery and also allow for shrinkage SAN to the trapezius may be very short. If the injury is of the graft substance. close to the border of the trapezius, sometimes it will be required to find viable fascicles distal to the injury into which grafts can be placed. If a lesion in continuity is dis- Results covered, external neurolysis is performed and standard Within a 32-year period, 156 patients with iatrogenic electrical and microscopic criteria are used to decide for SAN injuries resulting from cervical lymph node biopsy or against resection based on intraoperative NAP record- procedures were evaluated and found to require surgical ing across the lesion. repair. Of these 156 patients, 154 (99%) sustained injury In general, it is appropriate to resect soft tissue and during a biopsy procedure in Zone 1 of the PCT, where scar proximal and distal to the lesion and work toward the the nerve is most at risk for damage. Table 3 shows the

Fig. 3. Intraoperative photograph of a previous lymph node biopsy (yellow arrow) in Zone I of the PCT (right). The SAN is located at a point two-thirds of the distance (double arrows) starting from the clavicular insertion of the SCM to the mastoid process SCM origin. The incision for lymph node biopsy can be extended or turned into a Z-shaped incision by adding a limb either superiorly or inferiorly. Figure is available in color online only.

J Neurosurg Spine Volume 23 • October 2015 521

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC S. H. Park et al.

TABLE 3. Surgical procedures performed in 156 patients to repair extracranial SAN injury due to cervical lymph node biopsy Procedure No. of Patients Graft 82 End-to-end repair 41 Neurolysis 29 Neurotization (C2–3) 4 Total 156

who received local grafts compared with those who re- ceived sural grafts. Four patients underwent neurotization when the proxi- mal nerve trunk could not be identified. In this procedure, the contribution of the C-2 and/or C-3 nerves to the SCM was “swung down” to the trapezius and sewn to grafts that were attached to distal accessory branches. The results of neurotization were not as remarkable as for the other performed procedures, but still substantial. Two of the 4 patients who underwent neurotization experienced some degree of recovery. From a preoperative Grade 0, these 2 patients improved to Grades 2 and 3 after an average follow-up of 15 months.

Discussion Cervical lymph node biopsy is the most common cause of iatrogenic SAN injury, with an incidence of 3%–10%.1 The SANs superficial location in Zone I of the PCT, where Fig. 4. Intraoperative photographs. Upper: Exposure of the proximal most biopsy procedures are performed, places the nerve and distal stumps of a transected SAN. Lower: End-to-end repair of the at risk for injury and palsy, which results in winging of transected SAN is performed. Figure is available in color online only. the scapula and difficulty with shoulder abduction. Such injuries rarely undergo spontaneous regeneration, and generally require surgical intervention to repair the nerve. types of surgical procedures performed in each of the 156 In this study of 156 patients with SAN injuries during patients and Table 4 shows patient improvement by grade cervical lymph node biopsy, 154 received their injury in for those who were followed up for 12 months or more (n this area. The SAN emerges from the posterior or lateral = 152). border of the SCM through Zones I and II of the PCT. In Twenty-nine patients underwent neurolysis because the Zone I, the nerve is superficial and is embedded in cervi- nerve was in continuity and showed intraoperative elec- cal lymph nodes and fibroadipose tissue, making it sus- trical evidence of regeneration, such as a positive NAP ceptible to iatrogenic injury (Fig. 5). In some cases, the beyond the point of injury or trapezius contraction in re- nerve is also attached to the underside of the epidermis, sponse to SAN stimulation. Grade 3 or better functional making its manipulation necessary to access the tissues recovery was noted in 28 (97%) of these patients (Table of this zone. The SAN location makes it susceptible to 4). Nine of these patients reached Grade 5. All of these pa- stretching and laceration injuries, hence surgeons should tients experienced eventual dramatic improvement in arm use great caution when performing lymph node biopsies abduction and normalization of shoulder posture. in this area. The average length of the nerve at this point A total of 123 patients with lesions not in continuity or is 3.5 cm, after which it descends into the tissues in Zone with lesions in continuity but in which NAPs could not II before passing into the thoracic portion of the trapezoid be recorded underwent either end-to-end suture (41 pa- muscle. The SAN is generally absent in Zone III of the tients) or interposition graft repair (82 patients). The aver- PCT. A SAN injury should be suspected if a patient has age graft length was 3.81 cm. Of these 123 patients, 107 pain or stiffness in the shoulder and a history of an opera- (87%) achieved a Grade 3 or better functional recovery tion in the neck region, such as a lymph node biopsy or after surgery (Table 4). Of the 41 patients who underwent neck dissection.13 For suspected SAN injuries, examina- end-to-end repair, 37 (90%) recovered to Grade 3 or bet- tion of the function of the SAN should be performed to ter, which was slightly higher than the patients receiving make a timely diagnosis. The initial complaint of shoulder graft repair (85%). Most of the end-to-end repairs were pain in patients with SAN injuries may be misdiagnosed performed secondarily in more sharply transected SANs. as glenohumeral instability or muscle injury.11 Other diag- Postoperative outcomes did not differ between the patients noses that should be ruled out include cervical radiculopa-

522 J Neurosurg Spine Volume 23 • October 2015

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC Spinal accessory nerve injuries caused by lymph node biopsy

TABLE 4. Functional assessments of patients before and after surgical intervention* Neurolysis (n = 29) End-to-End Repair (n = 41) Local Graft (n = 82) Grade Preop Postop†‡ Preop Postop†§¶ Preop Postop†§** 5 0 9 0 6 0 11 4 0 12 0 25 0 50 3 0 7 0 6 0 9 2 0 1 0 3 0 6 1 7 0 8 1 9 4 0 22 0 33 0 73 2 * Grades in functionality for patients completing follow-up of 12 months or more (n = 152). † One hundred thirty-five (89%) of 152 patients who underwent neurolysis, end-to-end, or graft repair recovered to Grade 3 or higher. ‡ Twenty-eight (97%) of 29 patients who underwent neurolysis recovered to Grade 3 or higher. § One hundred seven (87%) of 123 patients who underwent end-to-end or local graft repair reached Grade 3 or higher. ¶ Thirty-seven (90%) of 41 patients who underwent end-to-end repair recovered to Grade 3 or higher. ** Seventy (85%) of 82 patients who underwent a local graft procedure recovered to Grade 3 or higher. thy, brachial plexopathy/neuritis, and long thoracic nerve and occur either immediately after the procedure or soon palsy, which cause some combination of shoulder girdle afterward. These symptoms include pain over the affected depression, limited shoulder motion, and cervical/scapular muscle; a feeling of heaviness or “dragging” in the arm; pain.5 Some physicians erroneously rule out SAN injury an inability to abduct the arm beyond the horizontal plane if the patient can shrug his or her shoulders, but the leva- and to elevate the arm for more than a few seconds; and tor scapula muscle normally assists the trapezius muscle finally, numbness over the angle of the jaw and around the in this function and can therefore produce a shrug on the ear, sometimes caused by associated damage to the greater affected side.1 auricular nerve.10,19 Presenting symptoms of SAN injuries are not subtle In this study, pain was a common complaint; most pa- tients experienced the onset of pain immediately or within 1 week of their lymph node biopsy. Arm pain and occipi- tal and posterior neck discomfort was occasionally associ- ated with this pain, and the pain was typically described as a constant, dull ache that was sometimes severe. These symptoms may be a result of chronic shoulder depression that creates tension on the shoulder joint or elements of the . As the muscle atrophies, the ligaments stretch and the musculoskeletal structure on the affected side shifts, putting additional stress on the surrounding muscles and structures. Scapular winging caused by accessory palsy can be differentiated from long thoracic palsy by using the “flip test.” In this test, the patient stands with his or her arm at the side, and the elbow is flexed to 90°. The examiner resists glenohumeral external rotation at the distal forearm while visually examining the scapula. A positive scapular flip sign occurs when the medial scapular border “flips” up, or lifts from the thoracic wall during resisted shoulder external rotation.5 The flip sign is also present in patients with a partial return of the upper trapezius, which acts as an upward rotator and elevator. Therefore, even when ac- tive, it cannot oppose the pull of the humeral external rota- tors. The middle and lower trapezius muscles also appear to stabilize the scapular medial border during resisted hu- meral external rotation; even when intact, the rhomboids are unable to properly perform this duty during the as- sessment. In contrast, pushing forward against resistance with the hand and arm and a fully extended elbow does not produce severe winging, which typically occurs in Fig. 5. Illustration of a cross-section of the neck showing the SAN is long thoracic palsy. covered solely by skin and fascia in the PCT. Lymph nodes are usually A review of 11 studies of patients who underwent sur- close by the SAN. Copyright Daniel H. Kim. Published with permission. gical management of their SAN injuries reported clini- Figure is available in color online only. cal recovery rates ranging from 50% to 100%.1 The 156

J Neurosurg Spine Volume 23 • October 2015 523

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC S. H. Park et al. patients selected for surgical intervention in the present fied in recent years to preserve the SAN, iatrogenic injuries study had complete or nearly complete loss of trapezius still occur in as many as 10% of cervical lymph node bi- muscle function, with denervation, as assessed by electro- opsy procedures. Surgeons who operate in this area should myography. Of the 152 patients who were followed for 12 have a thorough knowledge of the anatomy of the PCT and or more months postsurgery, 89% recovered to Grade 3 be aware of the risk of injury of the SAN, particularly in or better for functional activity. Neurolysis was performed Zone I. While some patients may improve with conser- in patients with lesions in continuity that were regenerat- vative management, surgical exploration is indicated for ing as shown by positive NAP recordings with favorable SAN injuries that do not demonstrate clinical or electrical functional outcomes in almost all of the cases. Lesions not signs of recovery for up to 3 months after the injury. As in continuity or lesions in continuity but in which NAPs shown in this study, appropriate and timely surgical explo- could not be recorded underwent either end-to-end su- ration and repair of iatrogenic damage to the SAN follow- ture or interposition graft repair with a high percentage ing lymph node biopsy can result in favorable outcomes. of favorable outcomes. End-to-end epineural repair was performed if an injured nerve was transected or unable to transmit an NAP and a short nerve gap after resection Acknowledgments existed. When the gap was too large for end-to-end repair We thank and acknowledge Sangwon Yeo for his expertise in or if the nerve was badly fibrosed or embedded in heavy medical illustration and Mary A. Hall, PhD, for her editorial sup- scar tissue, autografts were harvested and interfascicular port in the preparation of this manuscript. repair was performed. The donor nerves for these grafts were harvested from the local cervical sensory nerve or greater auricular nerve if it was suitable in caliber and References quality, or in other cases a sural nerve graft was used. 1. Boström D, Dahlin LB: Iatrogenic injury to the accessory Postoperative outcomes of the patients who received local nerve. Scand J Plast Reconstr Surg Hand Surg 41:82–87, 2007 grafts and those who received sural grafts did not differ. 2. Bradley PJ, Ferlito A, Silver CE, Takes RP, Woolgar JA, Stro- Although postoperative grades for the 4 patients who un- jan P, et al: Neck treatment and shoulder morbidity: still a derwent neurotization were lower relative to results of the challenge. Head Neck 33:1060–1067, 2011 other surgical procedures performed, 2 of the 4 patients 3. Camp SJ, Birch R: Injuries to the spinal accessory nerve: a experienced substantial recovery. In 34 nonoperative cases lesson to surgeons. J Bone Joint Surg Br 93:62–67, 2011 followed by the authors previously and reported in 2008, 4. Durazzo MD, Furlan JC, Teixeira GV, Friguglietti CU, Kulc- spontaneous recovery was poor if the initial grades were 2 sar MA, Magalhães RP, et al: Anatomic landmarks for local- or less by the LSUHSC system.12 Also, it should be noted ization of the spinal accessory nerve. Clin Anat 22:471–475, 2009 that full nerve transfer has been reported to be less effec- 5. Kelley MJ, Kane TE, Leggin BG: Spinal accessory nerve tive in pain relief or restoration of muscle function, and it palsy: associated signs and symptoms. J Orthop Sports should only be undertaken where repair or grafting cannot Phys Ther 38:78–86, 2008 be performed.3 6. Kessler J, Gray AT: Course of the spinal accessory nerve Whereas previous studies have reported SAN injuries relative to the brachial plexus. Reg Anesth Pain Med and outcomes,8,12 the current large case series study focus- 32:174–176, 2007 es exclusively on iatrogenic injuries due to cervical lymph 7. Kierner AC, Zelenka I, Heller S, Burian M: Surgical anatomy of the spinal accessory nerve and the trapezius branches of node biopsy for which spontaneous recovery was absent the cervical plexus. Arch Surg 135:1428–1431, 2000 and surgical management indicated. Early detection of 8. Kim DH, Cho YJ, Tiel RL, Kline DG: Surgical outcomes of SAN injury is key to monitoring the nerve and developing 111 spinal accessory nerve injuries. Neurosurgery 53:1106– a treatment plan. Because the nerve seldom spontaneously 1112, 1102–1103, 2003 regenerates, early surgery is recommended to avoid sig- 9. Lloyd S: Accessory nerve: anatomy and surgical identifica- nificant atrophy of the trapezius and long-term functional tion. J Laryngol Otol 121:1118–1125, 2007 deficits. In our experience, patients who undergo surgical 10. London J, London NJ, Kay SP: Iatrogenic accessory nerve repair within 3–6 months of injury have better functional injury. Ann R Coll Surg Engl 78:146–150, 1996 11. Martin RM, Fish DE: Scapular winging: anatomical review, outcomes than patients who present beyond 8–12 months diagnosis, and treatments. Curr Rev Musculoskelet Med of the injury. The difference in outcomes between these 1:1–11, 2008 patients is mostly related to the degree of muscle atrophy 12. Midha R: Accessory nerve, in Kim DH, Midha R, Murovic that occurs over time following the injury. This large series JA, et al. (eds): Kline and Hudson’s Nerve Injuries: Opera- of patients with SAN injuries following cervical lymph tive Results for Major Nerve Injuries, Entrapments, and node biopsy spans more than 3 decades and represents a Tumors, ed 2. Philadelphia: Saunders Elsevier, 2008, pp relatively common clinical problem. While this study war- 391–399 13. Nakamichi K, Tachibana S: Iatrogenic injury of the spinal rants greater awareness of the risk for SAN injury during accessory nerve. Results of repair. J Bone Joint Surg Am cervical lymph node biopsy, it also demonstrates the favor- 80:1616–1621, 1998 able functional outcomes that can be achieved following 14. Saman M, Etebari P, Pakdaman MN, Urken ML: Anatomic SAN repair if managed correctly. relationship between the spinal accessory nerve and the jugu- lar vein: a cadaveric study. Surg Radiol Anat 33:175–179, 2011 15. Schuller DE, Platz CE, Krause CJ: Spinal accessory lymph Conclusions nodes: a prospective study of metastatic involvement. Laryn- Although neck dissections in the PCT have been modi- goscope 88:439–450, 1978

524 J Neurosurg Spine Volume 23 • October 2015

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC Spinal accessory nerve injuries caused by lymph node biopsy

16. Soo KC, Guiloff RJ, Oh A, Della Rovere GQ, Westbury G: Innervation of the trapezius muscle: a study in patients under- Author Contributions going neck dissections. Head Neck 12:488–495, 1990 17. Symes A, Ellis H: Variations in the surface anatomy of the Conception and design: Kim, Kline. Acquisition of data: Kim, spinal accessory nerve in the posterior triangle. Surg Radiol Park. Analysis and interpretation of data: Kim, Park, Esquenazi. Anat 27:404–408, 2005 Drafting the article: all authors. Critically revising the article: 18. Tubbs RS, Stetler W, Louis RG Jr, Gupta AA, Loukas M, Kim, Esquenazi, Kline. Reviewed submitted version of manu- Kelly DR, et al: Surgical challenges associated with the mor- script: all authors. Approved the final version of the manuscript phology of the spinal accessory nerve in the posterior cervi- on behalf of all authors: Kim. Study supervision: Kim, Kline. cal triangle: functional or structural? J Neurosurg Spine 12:22–24, 2010 19. Williams WW, Twyman RS, Donell ST, Birch R: The poste- Correspondence rior triangle and the painful shoulder: spinal accessory nerve Daniel H. Kim, Department of Neurosurgery, Memorial injury. Ann R Coll Surg Engl 78:521–525, 1996 Hermann Health System, The University of Texas Health Science 20. Ying M, Ahuja A: Sonography of neck lymph nodes. Part I: Center, 6400 Fannin, Ste. 2150, Houston, TX 77030. email: normal lymph nodes. Clin Radiol 58:351–358, 2003 [email protected].

J Neurosurg Spine Volume 23 • October 2015 525

Unauthenticated | Downloaded 10/10/21 10:13 PM UTC